An early case of acquired amnesia in a child, following ECT, is reported by Geschwind (1974). The 10-year-old child had normal language prior to the ECT but subsequently developed both memory and language impairment. Geschwind attributed the language impairment to a retrograde amnesia which extended back into the period of language acquisition. An alternative explanation would be that the acquired amnesia and acquired dyphasia were distinct disorders. However, Geschwind argued that an acquired memory impairment in childhood may not only affect the acquisition of subsequent skills, as discussed above, but may also affect skills that have been previously acquired if the amnesia disrupts the storage, consolidation or subsequent retrieval of previously learnt material.
Ostergaard (1987) presented the first detailed analysis of a child with acquired amnesia. The 10-year-old child, C.C., developed amnesia following an episode of anoxia following water intoxication in the treatment of diabetic ketoacidosis, resulting in severe left hip-pocampal damage and at least partial right hippocampal damage. The child remained of normal intelligence but many memory skills were impaired. Ostergaard (1987) made the first attempt to determine the components of memory that might be affected and spared in such cases. He argued that the distinction between the skills that were impaired and those that were intact related to a procedural-declarative dimension, with declarative memory impaired and procedural memory intact.
Within declarative memory, both semantic and episodic memory was impaired. Knowledge of semantic memory for facts was impaired, as reflected by scores on the Information subtest of the WISC-R (Wechsler, 1974). The impairment in semantic memory also extended to vocabulary skills in tasks of lexical decision, semantic classification and verbal fluency. In lexical decision, accuracy was good, indicating development of lexical representations, but responses were very slow, suggesting access difficulties, and there was a significantly elevated priming effect of a related preceding word. In semantic classification, Ostergaard (1987) divided words by age of acquisition into three groups, 0-4 years, 4-8 years and over 8 years. He found a much sharper than normal temporal gradient to performance in terms of response speed, with words acquired early being responded to more quickly than those acquired later. He argued that this reflected a retrograde amnesia with a temporal gradient.
Impaired episodic memory was seen on immediate and delayed story recall, design recall and delayed free recall of word lists, although immediate free recall of words was within the normal range. Reading and spelling were also impaired and there was the suggestion of some retrograde loss, with teachers reporting normal reading prior to illness, yet reading and spelling ages 6 months after illness were found to be 9-14 months below chronological age at time of illness.
However, like cases of acquired amnesia in adulthood, C.C. had intact procedural memory for skills, such as learning a computer video-attack game. Like acquired amnesic patients, Ostergaard (1987) also showed that C.C. could learn and retain skills on the Gollin incomplete pictures tests, where pictures depicting an item become progressively more complete. Repeat testing 24 h later showed much improved skills for both C.C. and controls and the improvement was sustained on further retesting 8 days later. Follow-up at age 15 indicated a similar pattern, with a distinction between declarative and procedural memory skills. There was some limited progress in reading and spelling but much less than would be expected from a normal child.
Vargha-Khadem et al. (1992) reported, in an abstract, a case of acquired amnesia in a child, J.L., where they also found impaired performance in declarative memory but normal performance on procedural memory. J.L. developed amnesia after surgery for a cranio-pharyngioma. There was ventral diencephalic pathology encompassing the mammillary bodies. Some scholastic skills were nevertheless attained and reading developed to age level.
A further case of acquired amnesia is described in a 9-year-old child, T.C., who had acute encephalopathy, possibly as a result of herpes simplex encephalitis, resulting in diffuse cerebral injury (Wood et al., 1982, 1989). T.C. had severely impaired semantic memory in relation to facts about the world and severely impaired episodic memory impairment affecting both story recall and autobiographical day-to-day memory. However, Woods et al. (1987) argued that their case did not conform to the procedural-declarative distinction of the Ostergaard case, since T.C. made some scholastic progress through the school years, although a dense amnesia remained in both clinical and psychometric terms. However, Ostergaard & Squires (1990) argued, in response, that some scholastic progress would be expected in the absence of declarative memory on the basis of automated procedures or conceptual development. They also emphasized that the distinction they had proposed was based upon relative rather than absolute impairment, with studies of even acquired amnesia in adults discussing "differential susceptibility of these systems to amnesia" (Squire & Cohen, 1984). C.C.'s declarative skills were severely impaired and were much more impaired than procedural memory skills. Similarly, for T.C., declarative knowledge at age 20 remained severely impaired, with inability to report events of the last hour and memory for verbal and visual material "essentially absent", but there were some miminal skills, e.g. with some words recalled on repeated exposure with the Rey Auditory Verbal Learning task. Ostergaard & Squire (1990) argued that minimal residual declarative skills in C.C. or T.C. could enable some limited acquisition of skills contributing to scholastic progress, without difficulty for the Ostergaard (1987) proposal. They point out that in the cases of both C.C.
and T.C. scholastic progress in literacy and arithmetic was abnormally slow, although the Vargha-Khadem et al. (1992) case of J.L., mentioned above, illustrates that impairment in literacy is not a necessary concomitant of a declarative impairment.
Subsequent to this debate, Brainerd & Reyna (1992) argued that children's logical and mathematical skills and ability to make pragmatic inferences are not dependent upon memory in terms of reactivation of previously encoded traces. This would permit progress in these scholastic areas despite memory impairment. Wood et al. (1989) had reported above-average skills for T.C. in general problem-solving ability on the Porteus mazes, and a further pattern of this sort was confirmed by Broman et al. (1997), who reported the case of a child, M.S., with an acquired amnesia that followed respiratory arrest and consequent anoxic encephalopathy at the age of 8 years. MRI indicated loss of volume in bilateral hippocampal and medial temporal grey areas, with no evidence of amygdalar atrophy.
The child was followed up into adulthood and assessed in detail at the age of 28. Intelligence on Progressive Matrices was normal. In contrast, declarative memory was severely impaired. His profile on the Wechsler Intelligence Scales indicated that as an adult his weakest subtest scores were attained on Information, which assesses the factual general knowledge established in semantic memory, and on Vocabulary, which assesses the knowledge of words established in semantic memory. On further formal testing, he was impaired in episodic recall of words, stories and patterns. He was also impaired in learning paired associates and delayed recall of a route. His memory impairment extended to both antero-grade and retrograde loss as he was unable to remember any events more than 6-9 months prior to the anoxia.
Recognition memory for words, faces and doors on the WRMB (Warrington, 1984) and the Doors and People Test (Baddeley et al., 1994) was impaired but immediate recognition memory for patterns on Benton's Visual Retention Test was normal. Short-term memory, as assessed by digit span, was also normal. Consistent with the view of Brainerd & Reyna (1992), some mathematical and logical skills had become established, e.g. he developed algorithms to derive the multiplication tables he was unable to memorise by rote. With lengthy perseverance, he could solve puzzles like Rubic's cube and some computer games. On the Wechsler Intelligence Scales, Picture Completion, Object Assembly and Block Design were entirely normal, indicating intact skills of logical analysis, nonverbal reasoning and construction. He could also assemble items in real life, e.g. a canopy tent, and he could recite the tune and lyrics to the signature songs for his favourite television programmes. Thus, some automated and procedural memory skills had also been acquired. Within language, comprehension of syntax of varying complexity and comprehension of the sentences comprising the verbal ideational material taken from the BDAE (Goodglass & Kaplan, 1983), which incorporate adult logical complexity, were both at an average level for an adult. In contrast, receptive vocabulary was impaired and was at an average 9-year-old level; naming was also impaired to the expected level relative to his receptive vocabulary. Thus, he has a semantic representation anomia as discussed above. Reading and spelling remained at an 8-year-old level, hence whilst syntactic and logical reasoning skills in language had developed to an adult level, knowledge of word meanings and ability to read or spell words had remained severely impaired.
Broman et al. (1997) noted the difference in the patterns of deficit in the reported cases of acquired amnesia in childhood with respect to language and literacy development, with the degree of reading difficulty for M.S. being more extreme than that of C.C. (Ostergaard, 1987) or T.C. (Woods et al., 1982, 1989). Reviewing the identified lesions, he proposed that for those cases involving the hippocampus and its circuitry, language development and skills were also affected to a degree, which might vary dependent upon the degree of the amnesia. In cases like that of J.L (Vargha-Khadem et al., 1992), where the damage was to the diencephalic structures including the mammillary bodies, language and reading appear to develop normally, since reading was at the expected level for chronological age.
However, in contrast to this view, Vargha-Khadem et al. (1997) reported a further three cases of amnesic impairment in children following early hippocampal pathology. In each case they argued that language and literacy were in the low-average to average range and that episodic memory was much more significantly affected than semantic memory. Despite the bilateral hippocampal pathology, they argue for the development of semantic skills. These are the first cases of acquired amnesia in childhood in which it is argued that semantic skills are normal for both knowledge of words and factual knowledge of the world, and that some declarative skills are intact despite episodic impairment.
The first of Vargha-Khadem et al.'s (1997) three cases was a 14-year-old girl, Beth. After birth Beth remained without a heartbeat for 7-8 min before resuscitation. Memory difficulties were noticed on entrance into mainstream school. The second case, Jon, was a 19-year-old boy. He had been delivered prematurely at 26 weeks, had breathing difficulties and was in an incubator, on a ventilator, for 2 months. At the age of 4, he had two protracted seizures. Memory difficulties were noted by his parents at age 5.5 years. The third case was a 22-year-old, Kate, who for 3 days at the age of 9 had received a toxic dose of theophylline, a drug being given for her asthma. This led to respiratory arrest and loss of consciousness. Upon physical recovery, she displayed amnesia.
In support of competent semantic skills, Vargha-Khadem et al. (1997) discuss both the Information and Vocabulary subtests on the WISC-III (Wechsler, 1992) and predicted reading and spelling skills on the basis of IQ. They argue that there is a pronounced disparity between attainments on these measures and a severe amnesia for everyday life. In relation to the Information subtest, Beth and Jon attain normal scaled scores, although Kate has a scaled score of 6, indicative of a degree of impairment and equivalent to the level of performance taken to indicate impairment in the Ostergaard (1987) study. Scores on Vocabulary for all three children were within the normal range. Reading scores for all three children were commensurate with predictions from IQ, although for Beth, where intellectual skills were a standard deviation below average, these attainments were equivalently below average. Spelling was similarly at the predicted level for Beth and Kate, although this was more markedly impaired for Jon. Thus, whilst semantic memory is not intact for all skills in all three cases, there is clear evidence that it is has developed to a substantive degree.
For all three cases, immediate episodic memory for a word list was normal, as was digit span and Corsi span. Thus, short-term memory appeared normal. However, delayed recall of both verbal and nonverbal material was severely impaired. There was also parental complaint of severe difficulties with day-to-day memory to a degree that significantly affects their day-to-day abilities. This was confirmed for all three children by their very weak scores on the Rivermead Behavioural Memory Test (Vargha-Khadem et al., 1997). Thus, episodic memory is severely impaired and is much less well developed than semantic memory skills.
This inequality of sparing might result from independent storage, with the more significant impact upon episodic rather than semantic stores in these cases of bilateral hip-pocampal pathology, being similar to that seen in acquired amnesia. Another possibility that Vargha-Khadem et al. (1997) discuss is that, whilst episodic memory is impaired, it might be sufficiently preserved to enable the acquisition of knowledge to which there is repeated exposure in different contexts, and therefore enable context-free linguistic and factual knowledge. This would enable preservation of the theory of acquired amnesia for both adults and children, and would argue that both episodic memory and semantic memory are a single process mediated by the hippocampal system. However, Vargha-Khadem et al.'s (1997) proposal is that the underlying sensory memory functions of the perirhinal and entorhinal cortices may be sufficient to support context-free semantic memories, but not context-rich episodic memories, for which hippocampal circuitry is required. For both semantic and episodic memory to be affected severely, the hippocampi and underlying cortices would then require to be damaged.
Both Beth and Jon (Vargha-Khadem et al., 1997) had hypoxic-ischaemic episodes at or shortly after birth without showing any subsequent hard neurological signs. A further three similar cases are also presented by Gadian et al. (2000). In each of these five cases, the authors argue for the relative preservation of semantic memory over episodic memory. In each case, the Information and Vocabulary subtest scores on the WISC-III (Wechsler, 1992) are normal, and in each case basic reading is in line with IQ, episodic memory in terms of delayed story recall, delayed recall of listed words and delayed recall of the Rey figure is very poor, and day-to-day memory is impaired, based on both parental report and psychometric assessment. MRI scans indicate visible bilateral hippocampal atrophy in all cases and quantitative measures also indicate reduced grey matter in the putamen, with abnormality in the thalamus and midbrain.
Of further interest in the Gadian et al. (2000) paper is the parental comment, "Although his understanding of language is good, his use of language is often simplistic and he gropes for words". This suggests possible anomic difficulties in the use of language and, whilst name retrieval has not been explored in the cases of acquired amnesia discussed above, it has been found to be impaired in some of the cases of developmental amnesia outlined below (Temple, 1997a; Casalini et al., 1999).
A more detailed investigation of Jon, one of the hypoxic-ischaemic cases of Gadian et al. (2000), is given by Baddeley et al. (2001). Baddeley et al. (2001) established that, despite the very weak episodic recall skills, recognition memory for both visual and verbal material might nevertheless be normal. Recognition skills at a normal level were demonstrated on a range of tasks, including those which involved different speeds of presentation and those involving recognition after a 2 day delay. The only example of good recall came from material presented on a newsreel studied four times over a 2 day period, the conditions most like those involved in the acquisition of semantic memory. As Baddeley et al. (2001) note, Jon is an intelligent and highly motivated subject, so an above-average level of performance might have been expected. Thus, they do not argue that recognition is definitely normal but they do demonstrate convincingly that recognition is very significantly better than recall, even when scores are scaled to take account of the generally greater ease of recognition over recall. The enhancement of recognition over recall for Jon is of a degree that would not be typical of peer performance or the performance of adults with memory disorders.
Thus it would appear, from cases such as Jon's, that memory is being encoded and stored and that whilst the recollective process of episodic memory is impaired, this is not necessary for the acquisition of semantic knowledge or for recognition memory. This negates the view that semantic memory is simply the derivative of many episodes (e.g. Baddeley, 1997; Squire, 1992), a view derived from adult studies, where impaired episodic memory generally coexists with failure to update semantic memory, although Tulving (1972; Tulving & Markovitch, 1998) has argued otherwise for many years. It also indicates a more focal pattern of impairment in these cases of developmental amnesia than is typically seen in acquired amnesia. In acquired amnesia, recognition is usually, although not always, impaired (Aggleton & Brown, 1999; Holdstock et al., 2000) and, whilst semantic memory for facts acquired prior to injury is intact, there is difficulty in upgrading semantic information. In amnesic cases like Jon, recognition memory is good and semantic memory stores are upgraded, despite severe episodic memory impairment.
The idea of a developmental disorder being more focal than a similar acquired disorder in adulthood is interesting, given the view often argued that abnormalities in development have a pervasive and generalized impact because the system is adapting and formulating without prespecified functional architecture. Jon's case of amnesia from an early age demonstrated the potentially focal impact of a developmental memory impairment, with a pattern of performance entirely consistent with a modular view of the developing memory system.
Tulving (1985) defined episodic memory in a more focal way than it is often used now. He stressed the subjective experience of recollection in episodic memory. Thus, he distinguished between "remembering" and simply "knowing". Recognition memory is also argued to reflect these two processes of "remembering" and "knowing" (Tulving, 1985). Baddeley et al. (2001) argue that when Jon recognizes information, he "knows" he has seen it before but he does not "remember" the experience of having seen it before. In their terms, he lacks the ability to recollect the contextual detail which would be necessary for a "remember" response. Thus, the memory impairment is truly episodic in the original Tulving (1972, 1985) meaning of the term.
A further case of acquired amnesia, in which there are retrieval difficulties of a more unusual form, is described by Vargha-Khadem et al. (1994) in a 14-year-old boy, Neil. The amnesia followed successful treatment the preceding year with radiotherapy and chemotherapy for a tumour in the pineal region of the posterior third ventricle. Neil retained normal verbal intelligence, attaining scores of 111/109 on the verbal scale of the Wechsler Intelligence Scales. Although Performance IQ was significantly impaired, block design scores were normal. Retrograde memory was normal but episodic memory was impaired. Memory assessed on the Wechsler Memory Scales was very poor, with a MQ of 59 attained. Neil could copy a complex design well but delayed recall was severely impaired. With spoken verbal response, the verbal memory impairment appeared to be pervasive. There was also both agnosia and alexia. Although he could recognize familiar objects when they were in their customary place in his home, he was unable to recognize any of them when they were placed on a table directly in front of him. Yet, he could produce precise and intricate drawings of imagined objects and scenes and had intact and detailed visual memories for these items. Writing was also intact and he could produce accurate written responses for some information that he could not access in oral form, thereby indicating some ability to learn and retain new information. He was not always aware of his correct written responses. Thus, the recall is not linked to explicit awareness.
These abilities argued for separate stores and/or retrieval modalities for verbal material, one oral and one orthographic. His ability to use orthographic output included information taught in a verbal format at school but also included recall of day-to-day events. In the case of Jon (Baddeley et al., 2001), overt recall was problematic but recognition was much better. In the case of Neil (Vargha-Khadem et al., 1994), overt oral recall is problematic but written recall is much better. In both cases, it appears that memory is being encoded and stored but the recollective process of episodic memory is impaired and, without that episodic process, one may "know" rather than "remember", as in the case of Jon (Baddeley et al., 2001), or one may not have conscious awareness of "knowing", as in the case of Neil (Vargha-Khadem et al., 1994), despite explicit response accuracy with orthographic output.
Was this article helpful?
How Would You Like To Amaze People With Your Intelligence? Increase your IQ and get prepared to receive accolades in every sphere of life. Do you feel dejected every time your boss praises a colleague for an intelligent professional move? Do you want to become a crucial resource to your company?